�

�Q, Cash Flow and Investment: An

Econometric Critique

Christopher F. BaumBoston College

Clifford F. ThiesShenandoah University

August 12, 1997

We acknowledge the constructive remarks of two anonymous reviewers. The standard dis-claimer applies. Address correspondence to Clifford F. Thies, Shenandoah University, 1460University Drive, Winchester, VA 22601; 540-664-5450; cthies@su.edu.

Abstract

Q, Cash Flow and Investment: An Econometric Critique

The effects of measurement and speci�cation error on estimates of the Q

and cash �ow model of investment are investigated. Two sources of error are

considered: expensing of R&D expenditures and failing to identify that compo-

nent of cash �ow which relaxes �nancing constraints. We apply random-effects

and instrumental variables estimators to a model that addresses these sources of

error. We �nd that: (1) the capitalization of R&D strengthens the explanatory

power of the model; (2) expected and unexpected components of cash �ow have

different effects; and (3) the effects of Q are much more evident in �rms facing

low costs of external �nance.

2

1. Introduction

Q, Cash Flow and Investment:An Econometric Critique

The Q model of investment, for all of its analytical appeal, has achieved only

modest success in empirical research. Tobin�s Q, de�ned as the ratio of the

market value of the �rm to the replacement cost (or current cost) value of its

assets, can be shown to be a �sufficient statistic� for investment (Chirinko 1995).

Yet, beginning with the work of Fazzari, Hubbard and Petersen (1988), and

continuing through a growing and now international body of empirical research

(UK: Devereux and Schiantarelli (1990), Japan: Hoshi, Kashyap and Scharfstein

(1991), Germany: Elston (1993), and Canada: Schaller (1993)), Q has made only

a small contribution to the explanatory power of investment spending equations

that also include cash �ow or some other output-related variable.

The dogged search for the role played by Q in investment decisions is under-

standable. The Q model of investment, in addition to being soundly grounded

in theory, identi�es an explicit linkage between the real and �nancial sectors of

the economy. Furthermore, certain empirical regularities appear to be consistent

with the Q model: its performance appears to be stronger in subsamples con-

sisting of �rms that can be presumed not subject to liquidity constraints than

it does in subsamples of apparently liquidity-constrained �rms. This interplay

of Q and investment with capital market imperfections is itself intriguing, and

has led to a sizable literature (for a recent summary, see Schiantarelli, 1995,

pp.180-185).

Appearances, however, can be deceiving. Many prior investigators added

cash �ow variables or other measures of liquidity to the regression equation

implied by the Q model in an ad hoc manner. In contrast, Chirinko (1995)

3

derived the Q-and-cash �ow model from �rst principles, and determined that

if the additional costs faced by liquidity-constrained �rms take the form of a

higher cost of capital on funds raised, then Q already incorporates the effects of

capital market imperfection�it is a �sufficient statistic for investment� (p. 14),

and cash �ow need not be added to the model. However, he shows that if the

additional cost faced by liquidity-constrained �rms is in the form of a higher

upfront fee or �otation cost�a ��xed cost� component�then Q is no longer a

sufficient statistic, and a measure of liquidity (such as cash �ow) is required to

properly specify the investment model. Furthermore, a testable implication of

this analytical speci�cation is that a fundamental parameter of the model would

be larger for �rms facing more severe �nancing constraints. Chirinko found that

this relationship was typically contradicted in published research for a number

of countries and sample periods.

In this paper, we conduct an econometric investigation of the Q and cash

�ow model on �rm-level panel data, focusing on the sensitivity of the model

to measurement and speci�cation errors. We consider two aspects of potential

error: �rst, that intangible expenses (such as research and development) should

be capitalized, as their contribution to the value of the �rm is impounded in the

�rm�s stockmarket valuation, rather than expensed; and second, that it may be

inappropriate to treat anticipated and unanticipated cash �ow as equally effec-

tive in relaxing the liquidity constraints facing a �rm. We partition the sample

to account for differential costs of external �nance, and �nd that the model�s

ability to explain investment spending differs substantially between these sub-

samples.

The plan of the paper is as follows. In the next section, we brie�y survey

the performance of the Q model of investment, and consider how econometric

issues may explain the model�s apparent weaknesses. In Section 3, we describe

4

��

IK

� � � �= + ( 1) + +

I

Ka b Q c

F

K�

2. Measurement and Speci�cation Errors in Q Models

2.1. Measurement Error of Q and the Investment Rate

the data set which we employ in empirical analysis. Section 4 contains our

empirical �ndings, and Section 5 contains our conclusions and suggestions for

further research.

Many authors (as summarized in Chirinko, 1993) have noted that the Q model

has performed poorly in explaining macroeconomic investment. Problems in-

clude implausible parameter estimates as well as low explanatory power relative

to more traditional, accelerator-type models. Aggregation problems can cer-

tainly be at work (Abel and Blanchard 1986); so can measurement problems.

Micro-level data can address both sources of error; with respect to measurement

error, they may be used to increase the ratio of signal-to-noise in the data. How-

ever, Q research has been found to be quite sensitive to measurement problems

(Blundell et al., (1992), Perfect and Wiles (1994)). Perfect et al. (1995), for

example, obtained very different results when they used a single year�s estimate

of Q than when they used the average of three years� estimates. Klock, Thies

and Baum (1991) show that the standard imputation approach to �rm-level Q

measurement leads to serious discrepancies relative to direct observation of the

replacement cost value of assets and the traded or fair value of debt. They found

that not only is Q measured with error, but that the measurement errors are

likely to be correlated with �nancial ratios frequently employed in research.

Consider the following regression:

(2.1)

where is gross investment scaled by the capital stock, Q is Tobin�s Q, and

5

�

1

1

FK

=+

=1

+=

+a

��

� , b

� , c

�

� �

c/b

However, Chirinko (1993) does not �nd that measurement problems per se explain the poorperformance of the Q model of investment. As shown in Klock et al. (1991), that �nding maybe largely the result of Chirinko�s use of aggregate data (rather than �rm-level panel data).

is that part of cash �ow which serves to relax liquidity constraints, scaled by

the capital stock. As derived by Chirinko (1995, p.17),

(2.2)

where is the depreciation rate, is a parameter de�ning the rate at which

the cost of adjusting the capital stock rises, and is a parameter de�ning the

rate at which the �otation costs rise with the level of funds raised. Flotation

costs are assumed to be additively separable from adjustment costs and other

costs of production. A point estimate of may be recovered from a ratio of the

estimated coefficients ( ).

To the extent Q is measured with error�arising, for instance, from researchers�

inability to capture all aspects of the �rm�s valuation by the �nancial markets

in the numerator, and/or the appropriate replacement cost of its assets in the

denominator�its coefficient would be biased toward zero, and estimates of other

coefficients in the equation would be biased and inconsistent. If Q is the only

mismeasured explanatory variable, the attenuation of its coefficient value would

be in line with the common empirical result that the effect of Q on investment

is implausibly small. This measurement problem can be addressed satisfac-

torily via instrumental variables estimators if appropriate instruments can be

identi�ed.

However, the textbook presentation of �errors-in-variables� considers errors

of measurement which are uncorrelated with the true explanatory variables and

with the equation�s disturbance process. In the Q model, it is very likely that

an error of measurement for Q will be correlated with the included cash �ow

variable. Given that Q can be interpreted as the present value of future cash

6

2.2. Speci�cation of the Cash Flow Measure

�ow, a high correlation can be expected between the mismeasurement of Q and

current or lagged cash �ow variables.

A particularly signi�cant source of mismeasurement concerns off-balance

sheet assets, or intangible assets such as the value of a �rm�s technology as

developed by its expenditures on research and development. While accountants

will recognize a value for patents and other intangible assets purchased through

arms-length transactions, they will not recognize a value for internally-produced

intangibles. Yet, non-defense expenditures on research and development are

large, amounting to 1.9 percent of GDP in the United States in 1990, 3.0 per-

cent in Japan and 2.7 percent in Germany (Statistical Abstract of the United

States, 1993, p.598). In certain industries, these expenditures exceed investment

in plant and equipment. The value of intangible assets has been found to be

incorporated into the market�s valuation of �rms (Hall 1992). Chirinko (1993)

and Klock, Baum and Thies (1996) �nd that recognizing research and develop-

ment substantially improves performance of the Q model of investment. This

recognition involves rede�nition of investment spending and the capital stock to

incorporate spending on intangibles and the �stock� of intangibles, respectively.

A speci�cation issue with the Q-and-cash-�ow model arises in identifying that

part of cash �ow that serves to relax a �rm�s liquidity constraints. This has

been addressed in the line of research initiated by Fazzari et al. (1988) by

partitioning samples into two or more subsamples, including those considered

to be (more) liquidity-constrained and those considered to be non- (or less)

liquidity-constrained. Criteria employed have included dividend payout rates,

size as measured by sales or by total assets, and association with business groups

and/or banks (Schiantarelli, 1995, pp. 197-200). A larger coefficient on the cash

7

2

3

2

3

c/b

However, as Kaplan and Zingales (1996) point out, the relationship between investment-cash�ow sensitivity and �nancing constraints is ambiguous.

The analytical derivation of an investment model containing Q and cash �ow (e.g. thatof Chirinko, 1995) does not make the distinction between expected and unexpected cash �ow.However, since the estimated coefficients on these components may be chosen to be identical (orstatistically indistinguishable) by the data, we would argue that allowing for separate coefficients

�ow variable in the liquidity-constrained subsample is taken as con�rmation of

the model. Yet, as can be seen in equation (2.1) above, the parameter indicating

the cost of raising funds depends on the relationship between this coefficient and

the coefficient on the Q variable, and not simply on the coefficient on the cash

�ow variable itself. If an estimated model is consistent with this analytical

framework, the ratio ( ) should be greater for those �rms facing liquidity

constraints (or a higher cost of external �nance).

Recent advances in corporate �nance have examined how capital market im-

perfections, characterised as agency costs, can affect investment. Agency cost

problems may be minimized by a �rm�s ability to �bond� a part of cash �ow�for

instance, by committing cash �ow to debt service, and by the implicit commit-

ment arising from common dividend strategies. In this context, using total cash

�ow as a measure of liquidity, as is standard practice, is clearly misleading, as

the expected component of cash �ow may well be precommitted to meeting �-

nancial obligations. It would thus seem that the unexpected component of cash

�ow would be more effective in relaxing a �rm�s liquidity constraint than would

expected cash �ow. We focus in our empirical investigation on the distinction

between expected and unexpected components of �rms� cash �ow, and allow

those components to have different effects on investment spending. If the effects

of these components are indeed distinguishable, then failing to allow for the dis-

tinction will comprise potentially damaging speci�cation error�whether or not

unexpected cash �ow proves to be more important than expected cash �ow in

in�uencing investment. As in the model including a single cash �ow measure,

8

c/b

3. Data

Panel84

in implementing the model is consistent with the analytical framework.

the ratio ( ) of the coefficient on cash �ow to that on Tobin�s Q should be

greater for �rms facing a higher cost of external �nance; in our context, this

should hold with respect to components of cash �ow as well.

Our main data source is the data set, which consists of annual data

on 98 large U.S. manufacturing corporations over the period 1977-1983. These

data include �rms� disclosures of replacement cost values of inventory, plant and

equipment, cost-of-goods-sold and depreciation, as required �rst by the Securi-

ties and Exchange Commission Accounting Series Release 190 and then by the

Financial Accounting Standards Board Statement No. 33. On the liability side,

the data include market value �gures for traded debt and fair value imputations

for non-traded debt, using issue-speci�c data on coupon, term, conversion rights,

sinking fund and credit rating (see Thies and Sturrock 1987). We make use of

ten quantitative measures from this dataset: additions at cost, total assets at

replacement cost, gross cash �ow at replacement cost (net income plus depre-

ciation), our estimate of Tobin�s q, net sales, current assets, current liabilities,

�nancial leverage, interest expense, and an estimate of CAPM beta (generated

as described in Klock et al., 1996, p.390).

We supplement these data in two ways. From Hall (1990), we obtain data on

the �stock� of research and development capital. She develops these estimates

as the sum of past R&D expenditures, assuming a 15 percent depreciation rate.

Considering that �rms are only required to report research and development

expenditure if it is greater than one-half of one percent of sales, we assume ex-

penditure for non-reporting �rms�about 15 percent of the sample�is one-quarter

9

4

5

4

5

Value Line

Value Line.

We have reestimated all models removing the 166 �rm-year observations for which �R&Dstock� was imputed. There are no qualitative differences in the results from this restrictedsample.

We have also constructed estimates from a sample which excludes both R&D imputationsand expected cash �ow imputations. The results from this sample (which loses 188 �rm-yearobservations present in the full sample) are not qualitatively different from those presented inthe Tables.

of one percent of sales.

From 1976-1982 issues of , we obtain data on expected cash �ow:

speci�cally, the average of the investment service�s forecasts of cash �ow per

share times the number of outstanding shares and of net income plus deprecia-

tion, using �gures from the fourth quarters of the years preceding the years of

this study. In order to impute expected cash �ow for companies not monitored

by Value Line (about 10 percent of the sample) we impute values from a model

�t over those �rms for which we have complete data, making use of partial fore-

cast data where it exists. For those �rms lacking any forecast data, we assume

that their average error over the sample is the same as the average error for those

�rms with complete data. Table 1 presents some summary statistics. Interest-

ingly, expected cash �ow is systematically higher than actual cash �ow, with a

mean difference of 1.5 per cent of the capital stock (or about one-�fth of actual

cash �ow). Both expected and unexpected cash �ow measures are considerably

more variable than actual cash �ow. We have reestimated all models excluding

the 63 �rm-year observations for which expected cash �ow was imputed, and

�nd no qualitative differences arise from the restricted sample.

It might be argued that this measure of cash �ow�including net income rather

than EBIT�is not that most commonly used in the capital investment literature;

it is used in this context to provide consistency with the information on ex ante

forecasts available from Other studies have shown (e.g. Klock and

Thies, 1995) that varying de�nitions of cash �ow may have little effect on the

10

{ }

HIFLOAT

HIFLOAT

HIFLOAT

empirical �ndings of capital investment models. We have estimated the models

of this paper with both our measure of cash �ow (net income plus depreciation)

and the more commonly used measure (EBIT plus depreciation), not broken

into their expected and unexpected components, and have found no material

differences. Thus, the results below present only models estimated with our

original measure of cash �ow.

We utilized a speci�cation search to construct a scalar measure of the cost

of external �nance. Observations on �nancial leverage, interest expense as a

proportion of sales, the current ratio, the CAPM beta and total assets (as a

measure of size) were considered. This set of variables, and a number of sub-

sets, were analyzed in terms of the explanatory power of their �rst principal

component, and the sensibility of the elements of the related eigenvector. The

�rst three variables��nancial leverage, ratio of interest expense, and the current

ratio�comprised the �nal set chosen (descriptive statistics given in Table 1). The

�rst principal component of that set explains 46.7% of their total variation, with

positive loadings for leverage and interest expense, and a negative loading for the

current ratio. Thus, larger values of this proxy for the cost of external �nance

are generated by �rms with higher leverage, higher interest expense, and/or a

lower current ratio, making it a plausible measure of the �rm�s creditworthiness.

The sample is then divided into those �rm-years in which this proxy takes on

higher values than its median ( =1), and those for which it takes on

lower values ( =0). A given �rm is permitted to move between these

subsamples in subsequent years; thus, we do not classify �rms as facing rela-

tively high or low costs of external �nance, but rather classify �rm-years in that

way, allowing for deterioration or improvement in �rms� creditworthiness. De-

scriptive statistics for the = 0,1 subsamples are presented in Table

2.

11

Value Line

4. Empirical Findings

4.1. Random-Effects Estimates of the Standard Model

We do not consider pooled cross-section/time-series OLS estimates of the model,

as tests for the signi�cance of �rm and time effects indicate that models exclud-

ing those effects would be seriously misspeci�ed. Where practicable, we apply

the standard random-effects estimator, as it generalizes the �xed-effects speci�-

cation, attaining greater efficiency contingent on the orthogonality of regressors

and the error process. Hausman tests are used to establish the appropriateness

of the random-effects speci�cation.

In column (1) of Table 3 we present random-effects estimates of the standard

Q-and-cash-�ow model, comparable to equation (2.1) above, in which the whole

of cash �ow is presumed to relax �rms� liquidity constraints. Consistent with the

literature, the Q and cash �ow variables are lagged to avoid simultaneity bias

and to allow for time-to-build. The next several columns of this table investigate

the extent to which these problems are ameliorated, on the one hand, by the

distinction between expected and unexpected measures of cash �ow and, on the

other hand, by the recognition of intangible capital. In columns (2) and (4), cash

�ow is split into two components: expected cash �ow, equal to the

forecast, and unexpected cash �ow, equal to (actual) cash �ow less expected cash

�ow. In columns (3) and (4), gross investment is de�ned as gross investment

in plant and equipment plus R&D expense, and capital stock is de�ned as net

�xed assets plus net working capital plus R&D stock.

Comparing columns (2) and (4) to columns (1) and (3), we �nd that, when

cash �ow is separated into its expected and unexpected components, expected

cash �ow (ECF) enters the model with a larger coefficient than unexpected cash

�ow (UCF). In both cases, the difference between these estimates is statistically

12

{ }

HIFLOAT

HIFLOAT

HIFLOAT

HIFLOAT

HIFLOAT

HIFLOAT

4.2. Random-Effects Estimates by Cost of External Finance

signi�cant; expected and unexpected cash �ow measures do not have the same

effect on the investment rate. Although we might expect that the coefficients

on UCF should be larger than their ECF counterparts, from an econometric

standpoint the message is clear: common coefficients on the two components of

cash �ow represent a constraint not supported by the data.

Comparing columns (3) and (4) to columns (1) and (2), we �nd that when

research and development expenditures are capitalized, Q enters the model with

a coefficient that is highly signi�cant: indeed, of similar signi�cance as the

cash �ow variable(s). Versus the traditional model (excluding intangibles), the

explanatory power of the model roughly doubles.

We now separate the sample into subsamples based on , the proxy

measure of the cost of external �nance. =1 refers to those �rm-years

in which a high cost of external �nance was encountered. We no longer consider

models containing a single cash �ow variable, given the strong evidence against

that speci�cation. Table 4 presents results comparable to columns (2) and (4) of

Table 3: estimates of the model with R&D expensed (columns 1,2) or capitalized

(columns 3,4) for = 0,1 , respectively. Striking distinctions emerge

when the Tobin�s Q coefficient estimates are considered. For the =0

subsamples, Q is signi�cant, with sizable point estimates; for the

=1 subsamples, Q is far from signi�cant. Coefficient estimates for ECF and

UCF are signi�cant in all cases, but are distinguishable from one another only

in the =0 subsamples: that is, in those �rm-years corresponding to

the unconstrained Q model. This may well re�ect the strong response of a con-

strained �rm to available cash �ow�either predictable or unpredictable�whereas

an unconstrained �rm might be expected to follow its optimal investment plans,

13

2

2

HIFLOAT

HIFLOAT HIFLOAT

HIFLOAT R

HIFLOAT

R

with expected resources having the greater impact on those plans (as is borne

out by the estimates in columns 1 and 3).

The values of the ECF and UCF coefficient estimates are considerably larger

in the =1 subsamples, as would be expected from the analytics. The

estimated values of for both ECF and UCF are given at the foot of the Table.

Recall that Chirinko (1995) found that for liquidity-constrained �rms (those

facing a high cost of external �nance) should exceed the corresponding for un-

constrained �rms (a relation rarely supported in the empirical literature). In our

estimates, the value of for =1 exceeds that of for =0

for both ECF and UCF.

We also note that the model �ts considerably better, on the one hand, for

=0 (low cost of external �nance) �rms; their are higher, and

their standard errors of regression lower, than those for =1 �rms.

On the other hand, the same ranking may be made for capitalized R&D versus

expensed R&D: the models in columns (3) and (4) �t considerably better than

their counterparts in terms of . Several conclusions seem justi�ed: �rst, di-

vision of �rms� observations by their contemporaneous cost of external �nance

is meaningful, in that the Q-and-cash-�ow model �ts poorly for �liquidity con-

strained� �rms facing a high cost of external �nance. This is consistent with

the extensive literature on �rms� behavior subject to liquidity constraints (cf.

Schiantarelli, 1995): a �rm unable to capitalize on investment opportunities will

not invest, irregardless of the signal generated by its Q value, weakening the

causal link between Q and realized investment. Second, in the low cost of exter-

nal �nance subsample, the distinction between expected and unexpected cash

�ow is supported by the data. Third, in either subsample, the capitalization of

R&D signi�cantly improves the explanatory power of the model.

14

6

6

Value Line

HIFLOAT HIFLOAT

HIFLOAT

HIFLOAT

HIFLOAT

4.3. Instrumental Variables Estimates by Cost of External Finance

Since the cash �ow forecast is observed during the quarter preceding the yearto which it applies, contemporaneous expected cash �ow is a predetermined variable.

If the poor performance of Q is due to measurement error, an appropriate

econometric technique for estimating the investment equation is instrumental

variables. This involves a two-step procedure. In the �rst step, the poorly-

measured right-hand-side variables of the behavioral relation are regressed on a

set of identifying variables that are correlated with their true values and uncorre-

lated with the measurement error. In the second step, the predicted values from

the �rst step are used in estimating the behavioral relation. This methodology

may be used to deal with both measurement error and simultaneity bias.

Table 5 reports instrumental variables estimates of the Q-and-cash-�ow model,

both expensing and capitalizing R&D, by =0 and =1

subsamples. In these speci�cations, in which contemporaneous measures of Q

and unexpected cash �ow are incorporated into the model, the problems of en-

dogeneity and time-to-build as well as measurement error are all addressed by

the identi�cation of these right-hand-side variables by the following set of pre-

determined variables: lagged Q, lagged unexpected cash �ow, contemporaneous

and lagged expected cash �ow.

Columns 1 and 2 of Table 5 contains the estimates of the speci�cation in

which R&D is expensed, while columns 3 and 4 present the equivalent estimates

for capitalized R&D. We again �nd strikingly different estimates of the Q coef-

�cient: signi�cant for =0 subsamples (while insigni�cant in column

2, and marginally signi�cant in column 4), but much larger for capitalized R&D

than for expensed R&D. As in Table 4, the ECF and UCF coefficients may

clearly be distinguished in the =0 (low cost of external �nance) sub-

samples, but not in those for =1. In terms of explanatory power, we

15

2

5. Conclusions

R

HIFLOAT

HIFLOAT

again �nd that the model �ts much better in the HIFLOAT=0 subsample (in

terms of and standard error of regression), and markedly better for capital-

ized R&D than for expensed R&D. The ratios for both ECF and UCF exhibit

the relation predicted by Chirinko: larger values are associated with high cost

of external �nance ( =1) �rms� observations. This relation holds for

both expensed and capitalized R&D, but is more evident for the latter.

We cannot claim that the instrumental variables approach, allowing for con-

temporaneous values of Tobin�s Q and unexpected cash �ow, yields better es-

timates than those for the more traditional approach using lagged values. We

present the IV estimates to demonstrate the robustness of our �ndings: once

again, the model �ts better for =0 �rms, and for samples in which

R&D is capitalized. For �rms facing low costs of external �nance, expected and

unexpected cash �ow have differential effects on investment spending. In this

context, the IV estimates add support to our earlier �ndings, and add credence

to the conclusion that models �t over the entire sample would be misspeci�ed,

irregardless of estimation technique.

Our �rst and clearest �nding is that the Q model of investment is substantially

improved when �rm�s research and development expenditures are capitalized (as

opposed to expensed). As R&D expenditures grow relative to expenditures on

plant and equipment, and especially as R&D expenditures vary across industries,

it is increasingly important to take them into account in econometric work.

This �nding, of course, brings up the question if there are other sources of

intangible capital�e.g., advertising�that should be considered in this manner, as

did Klock et al. (1996). There is also a fundamental distinction which must be

faced by any attempt to capture intangibles� effects on pro�tability: should we

16

HIFLOAT

try to capture their effects by input-oriented measures, such as expenditures, or

proxy measures of output (such as patents, in the case of R&D)? We do not deal

further with this quandary here, but note that this distinction might not be so

important in our sample of large manufacturing �rms, which may be presumed

to invest in a diversi�ed portfolio of R&D projects, mitigating the difficulties of

proxying effectiveness by expenditure.

Our second �nding is that the Q model�s ability to explain investment expen-

ditures may be weakened substantially in the presence of high costs of external

�nance. Although we do not claim to have developed an ideal proxy for the cost

of external �nance, the proxy used here draws a clear distinction be-

tween �rms and years for which the Q-and-cash-�ow model �ts well, and those

for which it does not. Although analytical �ndings suggest that the Q model

(properly augmented by measures of liquidity) should be effective for all �rms,

we do not �nd strong support for the model among data evidencing a high cost

of external �nance.

A third �nding is that the distinction between expected cash �ow (which

might not re�ect �liquidity� due to precommitments) and unexpected cash �ow

appears important in those �rm-years where the Q-and-cash-�ow model �ts

adequately. From an econometric standpoint, the treatment of cash �ow as a

single, indivisible entity in the context of a Q model of investment spending

would appear to be clearly inappropriate.

17

Notes: Statistics are calculated from annual data, 1977-1983, for 98 large U.S. man-ufacturing corporations (a total of 686 observations). (I/K)*, Q*, (CF/K)*, (ECF/K)*and (UCF/K)* refer to those measures inclusive of R&D expenditures and R&D capitalstock estimates.

Table 1. Descriptive Statistics from the Panel84 Sample

Mean Standard Error Minimum MaximumI/K 0.069 0.044 0.004 0.541

(I/K)* 0.080 0.045 0.009 0.483Q 0.916 0.416 0.375 4.078Q* 0.825 0.349 0.361 3.414

CF/K 0.079 0.041 -0.199 0.336ECF/K 0.095 0.079 -0.204 1.494UCF/K -0.015 0.075 -1.395 0.264(CF/K)* 0.072 0.036 -0.177 0.291

(ECF/K)* 0.085 0.063 -0.175 1.089(UCF/K)* -0.013 0.061 -1.017 0.229Leverage 0.591 0.832 0.000 9.605

IntExp%Sales 1.951 1.846 0.000 22.158Cur. Ratio, % 27.845 12.905 -26.721 67.466

18

HIFLOAT HIFLOAT

HIFLOATNotes: Statistics are calculated from annual data, 1978-1983, for 98 large U.S.

manufacturing corporations. The variable divides the full sample on thebasis of the median of a principal component into subsamples containing 343 �rm-years.See notes to Table 1.

Table 2. Descriptive Statistics for Cost of External Finance Subsamples

=0 =1Mean Std.Error Mean Std. Error

I/K 0.072 0.040 0.065 0.047(I/K)* 0.086 0.044 0.073 0.045

Q 0.995 0.513 0.806 0.253Q* 0.889 0.434 0.736 0.045

CF/K 0.097 0.034 0.065 0.037ECF/K 0.115 0.110 0.065 0.037UCF/K -0.019 0.108 -0.013 0.035(CF/K)* 0.087 0.030 0.059 0.033

(ECF/K)* 0.103 0.087 0.071 0.031(UCF/K)* -0.016 0.086 -0.012 0.031

19

t

t t

t t

t t

�

� �

� �

� �

1

1 1

1 1

1 1

2

[ ]

[ ]

[ ]

[ ]

Pr CF�RPr �

Pr CF�

� �Pr �

Table 3. Random Effects Estimates of the Investment Rate

(1) (2) (3) (4)R&D Expensed Expensed Capitalized Capitalized

Constant 0.0380 0.0331 0.0360 0.0273(0.0072) (0.0074) (0.0074) (0.0077)

Q 0.0090 0.0084 0.0246 0.0249(0.0062) (0.0061) (0.0071) (0.0069)

CF /K 0.2649 0.2911(0.0596) (0.0642)

ECF /K 0.3060 0.3642(0.0613) (0.0667)

UCF /K 0.2506 0.2717(0.0595) (0.0637)

S.E.R. 0.0425 0.0422 0.0412 0.04080.0129 0.0005

0.3897 0.3795 0.4638 0.45130.125 0.139 0.199 0.2250.006 0.003 0.010 0.003

20

Notes: Estimates are calculated from annual data, 1978-1983, for 98 large U.S.manufacturing corporations (a total of 588 observations) and include a set of year dum-mies. Standard errors are given in parentheses. S.E.R. is the standard error of regres-sion. is the tail probability for the test of equality between the coefficientson expected and unexpected cash �ow. is the estimated GLS weight applied to thebetween-group variation, where =0 corresponds to OLS, and =1 corresponds to the�xed-effects estimator. is the tail probability for the F-test that time effects areinsigni�cant.

�

� �

� �

1

1 1

1 1

2

[ ]

[ ]

[ ]

[ ]

t

t t

t t

ECF

UCF

ECF

UCF

HIFLOAT

Pr CF�RPr �

Pr CF�

� �Pr �

Table 4. Random-Effects Estimates by Cost of External Finance

(1) (2) (3) (4)R&D Expensed Capitalized

0 1 0 1constant 0.0260 0.0304 0.0205 0.0326

(0.0091) (0.0123) (0.0100) (0.0126)Q 0.0092 0.0071 0.0309 0.0122

(0.0058) (0.0127) (0.0071) (0.0144)ECF /K 0.2888 0.4493 0.3213 0.5089

(0.0846) (0.1007) (0.1011) (0.1094)UCF /K 0.2227 0.3075 0.2219 0.3553

(0.0824) (0.0926) (0.0977) (0.0952)S.E.R. 0.0343 0.0444 0.0351 0.0416

0.002 0.152 0.000 0.1450.088 0.113 0.284 0.2090.192 0.116 0.309 0.1400.000 0.814 0.000 0.81031.5 43.6 7.2 29.224.3 63.7 10.4 41.8

21

Notes: Estimates are calculated from annual data, 1978-1983, for 98 large U.S.manufacturing corporations (a total of 588 observations), and include a set of yeardummies. Standard errors are given in parentheses. S.E.R. is the standard error ofregression. is the tail probability for the test of equality between the coefficientson expected and unexpected cash �ow. is the estimated GLS weight applied to thebetween-group variation, where =0 corresponds to OLS, and =1 corresponds to the�xed-effects estimator. is the tail probability for the F-test that time effects areinsigni�cant. is the ratio of the point estimates of the ECF/K and Q coefficients;

is the ratio of the point estimates of the UCF/K and Q coefficients.

� � � � �

2

1 1 1 1 1

[ ]

[ ]

[ ]

[ ]

t

t t

t t

ECF

UCF

ECF

UCF

HIFLOAT

Pr CF�RPr �

Pr CF�

� �Pr �

Table 5. Instrumental Variables Estimates by Cost of External Finance

(1) (2) (3) (4)R&D Expensed Capitalized

0 1 0 1constant 0.0231 0.0360 0.0187 0.0381

(0.0063) (0.0079) (0.0066) (0.0079)Q 0.0082 0.0084 0.0325 0.0165

(0.0040) (0.0093) (0.0046) (0.0100)ECF /K 0.3453 0.4045 0.3548 0.4366

(0.0667) (0.0791) (0.0740) (0.0826)UCF /K 0.1939 0.5896 0.1474 0.6729

(0.0719) (0.2273) (0.0809) (0.2304)S.E.R. 0.0751 0.0315 0.0243 0.0295

0.000 0.346 0.000 0.2970.149 0.110 0.339 0.2050.186 0.099 0.294 0.1230.000 0.349 0.000 0.29723.6 70.1 4.5 40.941.8 48.2 10.9 26.5

22

Notes: Estimates are calculated from annual data, 1978-1983, for 98 large U.S.manufacturing corporations (a total of 588 observations), and include a set of yeardummies. Instruments include Q , ECF/K, ECF /K , UCF /K and the yeardummies. Standard errors are given in parentheses. S.E.R. is the standard error ofregression. is the tail probability for the test of equality between the coefficientson expected and unexpected cash �ow. is the estimated GLS weight applied to thebetween-group variation, where =0 corresponds to OLS, and =1 corresponds to the�xed-effects estimator. is the tail probability for the F-test that time effects areinsigni�cant. is the ratio of the point estimates of the ECF/K and Q coefficients;

is the ratio of the point estimates of the UCF/K and Q coefficients.

References

Econometrica

Journal of Econometrics

Journal of Economic Dynamics and Control

Asymmetric Information, Corporate Finance, and Investment

Brookings Papers on Eco-nomic Activity

Quarterly Journal of Economics

Financial Review

[1] A.B. Abel and O.J. Blanchard, 1986. �The Present Value of Pro�ts andCyclical Movements in Investment.� 54, 249-273.

[2] R. Blundell, S. Bond, M. Devereux and F. Schiantarelli, 1992. �Does Qmatter for Investment? Some Evidence from a Panel of U.K. Companies.�

51, 233-257.

[3] Robert S. Chirinko, 1993. �Multiple Capital Inputs, Q, and InvestmentSpending.� 17, 907-938.

[4] Robert S. Chirinko, 1995. �Finance Constraints, Liquidity and InvestmentSpending: Cross-Country Evidence.� Unpublished working paper, EmoryUniversity.

[5] Michael Devereux and Fabio Schiantarelli, 1990. in R. Glenn Hubbard, ed.,. Chicago:

University of Chicago Press. 279-306.

[6] Julie Ann Elston, 1993. �Firm Ownership Structure and Investment: Ev-idence from German Manufacturing 1968-1984.� Unpublished working pa-per, Wissenschaftszentrum Berlin.

[7] Steven M. Fazzari, R. Glenn Hubbard and Bruce C. Petersen, 1988. �Fi-nancing Constraints and Corporate Investment.�

144-195.

[8] B.H. Hall, 1990. �The Manufacturing Sector Master File: 1959-1987�. Un-published working paper, National Bureau of Economic Research.

[9] B.H. Hall, 1992. �The Value of Intangible Corporate Assets: An Empiri-cal Study of the Components of Tobin�s Q.� Unpublished working paper,National Bureau of Economic Research.

[10] Takeo Hoshi, Anil Kashyap and David Scharfstein, 1991. �CorporateStructure, Liquidity, and Investment: Evidence from Japanese IndustrialGroups.� 106, 33-60.

[11] Steven Kaplan and Luigi Zingales, 1996. �Do Investment-Cash Flow Sensi-tivities Provide Useful Measures of Financing Constraints?� Unpublishedworking paper, Graduate School of Business, University of Chicago.

[12] Mark Klock and Clifford F. Thies, 1995. �A Test of Stulz�s OverinvestmentHypothesis.� 30:3, 387-398.

23

Journal of Economics andBusiness

Journal of Economics andBusiness

Journal of Banking and Finance

Journal of Empirical Finance

Canadian Journal of Economics

Is BankLending Important for the Transmission of Monetary Policy?

Journal of Accounting, Auditing and Finance

[13] Mark Klock, Clifford F. Thies and Christopher F. Baum, 1991. �Tobin�s qand Measurement Error: Caveat Investigator.�

43, 241-252.

[14] Mark Klock, Christopher F. Baum and Clifford F. Thies, 1996. �Tobin�sQ, Intangible Capital, and Financial Policy.�

, 48:387-400.

[15] Steven B. Perfect, David R. Peterson and Pamela P. Peterson, 1995. �Self-tender Offers: the effects of Free Cash Flow, Cash-Flow Signalling, and theMeasurement of Tobin�s q.� 19, 1005-1023.

[16] S. Perfect and K. Wiles, 1994. �Alternative constructions of Tobin�s q: anempirical comparison.� 1, 313-341.

[17] Huntley Schaller, 1993. �Asymmetric Information, Liquidity Constraints,and Canadian Invesment.� 26:3.

[18] Fabio Schiantarelli, 1995. �Financial Constraints and Investment: A Criti-cal Review of Methodological Issues and International Evidence,�in

J. Peek andE. Rosengren, eds., Federal Reserve Bank of Boston Conference Series No.39.

[19] Clifford F. Thies and Thomas Sturrock, 1987. �What did In�ation Account-ing Tell Us.� 2, 375-391.

24